Research On 3D Spatial Data Models And Refined 3D Automatic Modeling Methods For Underground Pipeline

As a vital component of the urban infrastructure, urban underground pipeline network (UUPN) is the paramount carrier for energy distribution, mass transport, information exchange, drainage for flood reduction and waste disposal. With the ever-increasing city population and unprecedented pressure from the city activities, the UUPN-an important underground space object-is becoming the focus of the important urbanization endeavor of China. The UUPN increasingly renders complexity, overlapping three-dimensional (3D) property, unpredictability and high-level difficulty during construction and needs significant investment. Furthermore, it is not easy to modify once the construction starts. Consequently, it is essential for UUPN implementation to achieve precise and detailed location management goals, to secure monitoring and management integration, to be highly efficient in emergency response and mitigation as well as comprehensive and extensive information sharing. Nevertheless, due to the non-uniformity of the current UPN-oriented 3D spatial data model and relevant standards, data information content, spatial representation and analysis capacity are insufficient. In addition, existing 3D models lack comprehensive refined modeling methods as well as efficient indoor/outdoor integration. Subsequently, it is challenging to reach a consensus in representation, understanding and sharing of UPN information resources, causing the established UPN 3D GIS being "an isolated information island". Together with low automation level of 3D modeling and difficult updating, all of above-described insufficiency can hardly support various UPN knowledge models in this profession.To address the above problems, this dissertation proposed an integrated geometric-temporal-topological-semantic 3D UPN spatial data model, leveraging the establishment of Kunming UUPN Information Application and Sharing Platform, using UUPN entity as research object based on a theory and application research methodology. A 3D modeling approach, with a high-level automation, refined, convenient updating and maintenance, is pursued based on this model. The contributions of this dissertation are summarized as follows:(1) After a comprehensive analysis of the requirements and contents of intelligent pipeline network-oriented UPN management, the dissertation proposed integrated 3D UPN spatial data models. The proposed models focused on the UPN object and consist of semantic data model, geometric data model and temporal data model components. Through the mapping and transformation from the real-world UPN to UPN ontology system to UPN object class library and then to UPN spatial-temporal objects, the models integrated geometric features, time features, topology features, and semantic features of the UPN entities. The modeling system results in efficient computing implementation of the models as well as lossless exchange and execution the sematic information of the UPN.(2) The dissertation proposed an innovation approach to automate and fully establish UPN fine models based on a hybrid CSG+Sweep+Boolean operation. This approach categorized the various UPN entities into 4 types:the abstract irregularly-shaped pipe points, the size/structure-driven pipe points, the topology-connection-relationship driven pipe points and the pipe segments. Different modeling methods have been employed for different category of entities. The dissertation also designed a data structure composed of three (3) basic geometry elements and seven (7) entity objects. Methods of inside-outside-of-well-chamber expression have also been researched in the dissertation.(3) This dissertation also studied online updating methodology of 3D UPN models. Facing the challenges of lengthy 3D UPN model establishing and thus not suitable to meet emergency application needs, the dissertation adopted a service-oriented architecture (SOA) and used XML for data exchange between the servers and clients. In such, after two-dimensional (2D) UPN change, a localized automated simultaneous online 3D UPN model data updating workflow was implemented, filling a blank of the profession in domestic China.(4) Based on the above research findings, the dissertation elaborated on the technical strategy of 3D system implementation and the efficiency of the resulting models, using Kunming UPN Information Application and Sharing Platform as prototype. Feasibility, effectiveness, and application value of the research were verified and validated with two real-world use cases.